Faculty of health sciences School of Public Health 0204521g dgresie@yahoo.co.uk;dgresie@hotmail.com / Ethylene oxide is used widely in hospitals as a gaseous sterilant for heat-sensitive medical items, surgical instruments and other objects and fluids that come into contact with biological tissues. Although ethylene oxide is recognised as a reproductive toxicant in humans, so far few studies have been carried out to investigate the association between exposure to ethylene oxide and the occurrence of adverse reproductive outcomes (Hemminki et al 1982 and 1983; Rowland et al, 1996; Yakubova et al, 1976). The results of these studies suggested that ethylene oxide is capable of causing reproductive dysfunction and that further research is needed in order to understand its effects on reproductive health. This study investigated the association between exposure to ethylene oxide during pregnancy and adverse reproductive outcome in women sterilising staff working in sterilising units using ethylene oxide in Gauteng province, South Africa. The study had the following objectives: 1) to describe the extent and nature of ethylene oxide use in sterilising units operational in medical facilities in Gauteng; 2) to assess the current exposure to ethylene oxide in sterilising units in Gauteng; 3) to collect information on the last recognised pregnancy using a questionnaire; 4) to assess the validity of the information on the evolution and outcome of the last recognised pregnancy collected by the means of the questionnaire; 5) to assess the association between occupational exposure to ethylene oxide during pregnancy and adverse reproductive outcome. The study population was represented by singleton pregnancies that: 1) occurred in women currently working in sterilising units using ethylene oxide in Gauteng province, South Africa; 2) were the last recognised pregnancy occurring in these women after the 1st January 1992; 3) occurred while the mother was employed. The adverse reproductive outcome was defined as the occurrence of any the following: spontaneous abortion, still birth, pregnancy loss (spontaneous abortion or still birth), low birth weight and combined adverse reproductive outcome (spontaneous abortion, still birth or low birth weight). The study enrolled 68.8% of the medical facilities in Gauteng that were using ethylene oxide to sterilise medical equipment. The majority of the employees working in the sterilising units included in the study were women (96.6%) and they were employed in one of the following jobs: technician (operator), instrument packer and cleaner. xiii Most of the sterilising units participating in the study used ethylene oxide sterilisation daily and only 15.4% of them reported that the employees operating the ethylene oxide steriliser used protective clothing. Recorded levels of ethylene oxide were provided by 46.2% of the sterilising units; they were all bellow 0.25 ppm (the South African long-term exposure limit for occupational exposure to ethylene oxide is 5 ppm). Changes in ethylene oxide sterilisation equipment and or technology were reported by 42.3% of the sterilising units and they were all engineering control measures aimed at reducing exposure to ethylene oxide. Measurements of the current levels of ethylene oxide were performed at the time of the study by the National Institute for Occupational Health using hydrobromic acid-coated petroleum charcoal tubes connected to calibrated Gilian pumps through which air containing ethylene oxide was drawn. The samples were analysed by the Analytical Services of the National Institute for Occupational Health. A total of 418 samples were collected (100 blank samples, 97 personal samples and 221 static samples). Quality control was ensured by the following methods: 1) verification by an Approved Inspection Authority; 2) collection of duplicate samples; 3) collection of blank samples. These measurements showed that exposure to ethylene oxide still occurred in sterilising units (ethylene oxide was detected in 9 out of the 10 public hospitals) and that the employees most exposed are the ones working with the ethylene oxide steriliser (technician or operator). There were 113 women working in the sterilising units enrolled in the study who had been pregnant after the 1st January 1992; 109 of them agreed to participate in the study and to complete the questionnaire. Information on exposure to ethylene oxide during pregnancy was obtained from three sources: walk-through survey, questionnaire-collected data and measurements of the levels of ethylene oxide in sterilising units at the time of the study. Information on the evolution and outcome of these pregnancies was gathered from the mother using a questionnaire. The questionnaire collected demographic data, reproductive history, medical data, risk factors for the adverse reproductive outcome (environmental and occupational exposures, lifestyle), and data regarding the evolution and outcome of the last recognised pregnancy. The questionnaire also collected detailed information on the job held at the time of the last recognised pregnancy (if the woman was working with ethylene oxide, she was asked to provide a complete list of daily tasks she was performing). Prior to administration, the questionnaire was tested on a small sample of working women. xiv The validity of the questionnaire-collected information on the evolution and outcome of the last recognised pregnancy was assessed by comparing this information against medical records (considered the “gold standard”). The assessment showed that mothers’ recall was accurate for the following variables: medical facility were the pregnancy was recorded, date of the reproductive event, gestation length, vital status of the newborn, number of foetuses, child gender, disease/medical problems during pregnancy and treatment received during pregnancy. There was an error in the mothers’ reporting of the birth weight of their babies. The possible misclassification of outcome resulting from this error was shown to be nondifferential (the proportion of subjects misclassified on outcome did not depend on exposure). Therefore, this misclassification could bias the effect estimate towards the null value or it could not produce any bias at all. The analysis carried out to detect possible associations between exposure to ethylene oxide and adverse reproductive outcomes included 98 of the initial 109 pregnancies on which information had been collected (11 pregnancies were excluded from the analysis for the following reasons: 2 were multiple pregnancies, 4 were conceived before 1st January 1992 and 5 were conceived while the mother was not employed). Amongst the 98 singleton pregnancies included in the analysis, 19 were classified as exposed and 79 as unexposed to ethylene oxide. The relative risk for spontaneous abortion was RR=16.63 (95%CI=1.97-140.42; p=0.004), for stillbirths RR=3.47 (95%CI=0.63-19.01; p=0.18), for pregnancy loss RR=6.24 (95%CI=1.95- 19.93; p=0.003), for low birth weight RR=0.61 (95%CI=0.09-4.30; p=0.51) and for combined adverse reproductive outcome RR=2.09 (95%CI=1.00-4.36; p=0.06). No confounders were detected for any of the associations between exposure to ethylene oxide and the adverse reproductive outcomes under study. For the association between exposure to ethylene oxide and combined adverse reproductive outcome the analysis detected three effect modifiers: paternal age (father aged 40 or older at conception), passive smoking and maternal age (mother aged 35 or older at conception). In conclusion, this study, the first in South Africa on ethylene oxide exposure and adverse reproductive outcomes, confirmed the widespread use of ethylene oxide, exposure to this agent in public sector hospitals and associations between exposure to ethylene oxide and spontaneous abortion and between exposure to ethylene oxide and pregnancy loss (either spontaneous abortion or stillbirth). xv Moreover, the study provided data on reproductive outcomes in employed women (on which scant data are available in South Africa) and added information on the validity of selfreported pregnancy data relative to medical records. The findings of the study support the conclusions of the previous studies that had suggested that exposure to ethylene oxide during pregnancy could lead to adverse reproductive outcomes. The study detected no associations between exposure to ethylene oxide and stillbirth, low birth weight or between exposure to ethylene oxide and combined adverse reproductive outcome.